Research On Precise Assembly And Measurement Of Three-dimensional Errors For Miniature Parts

Micro electromechanical systems(MEMS), which are composed of different parts, are demand for complex steps and high precision of assembly. At present in home, the assembly processes of most MEMS products with the feature size of several millimeters are still finished by experienced workers under microscopes. These result in not only labor intensity and low efficiency, but also the low consistency of products. At the process of error measurement of assembled parts, the operators usually carry parts from manufacturing line to high precision and expensive coordinate measuring machine. It is random that every chosen reference and measurement because of no standard measure procedures. Due to no relationship between assembly and measurement, it will spend longer time on the whole production cycle. At the meantime, the assembled MEMS products are easy to be damaged during transport.With the features of non-contact and high accuracy, the machine vision measurement technique works well for the semiautomatic and automatic precision assembly and measurement task for miniature devices. Based on the machine vision measurement technique and the request of assembly and measurement process, an automatic assembly and measurement system for precise miniature devices is introduced. The system is divided into the machine vision based measurement module, the manipulation module, the assembly working platform module and the human-machine interface module.In the process of assembly, a control method of "look-move" is introduced. It means that the results measured by the vision system guide the operation of manipulator. By employing the mixed forms of lighting, it’s convenient to extract feature from the background. The vacuum gripper with an integrated force sensor is designed to ensure the flexible and reliable clamping of different fragile devices. A mathematical model of assembly system is developed, so through the coordinate transformation method different modules can collaborate in a same coordinate. The procedures of assembly are properly planned to improve efficiency. In the process of measurement, a measuring method of two vision systems which are placed orthogonally is introduced. A standard template’s two sidelines processed in precision is used to determine the relative position of two vision systems. Then the two vision systems can work in collaboration to measure the devices with three-dimensional feature. The method has advantages of being simple, fast and accurate.The extraction and recognition of devices feature are realized through the function in the open source computer vision library. Different locating methods should be taken for the different devices characteristics.An assembly experiment is performed with the control programs of assembly and measurement process edited by the thought of task-strategy-behavior. The experimental results which are tested by the3-D assembling error measurement function of the developed system and high precision coordinate measuring machine show that the process of assembly and measurement can be finished well by the developed system introduced in this paper. At the meantime, the production cycle is shorten and the assembly efficiency and yield are improved.